Comparing one column value to all columns in linux enviroment

问题

So I have two files , one VCF that looks like

88  Chr1    25  C   -   3   2   1   1
88  Chr1    88  A   T   7   2   1   1
88  Chr1    92  A   C   16  4   1   1

and another with genes that looks like

GENEID  Start END
GENE_ID 11 155
GENE_ID 165 999

I want a script that looks if there is a gene position (3rd column of VCF file) within the range of second and third position of the second file and then to print it out.

What I did so far was to join the files and do

awk '{if (3>$12 && $3< $13) print }' > out

What I did only compares current rows of joined files (it only prints if the value is in the same row), how can I make it compare all rows of column 3 to all rows of column 12 and 13?

Best, Serg

回答1:

I hope to help (EDIT i change the code for more efficient algorithm)

gawk '
  #read input.genes and create list of limits (min, max)
  NR == FNR {
    #without header in input
    if(NR>1) {
      for(i=$2; i<=$3; i++){
        limits[i]=limits[i]","$2"-"$3;
      }
    };
    next
  }
  #read input.vcf, if column 3 is range of limits then print
  {
    if($3 in limits){
      print $0, "between("limits[$3]")"
    }
  }' input.genes input.vcf

you get:

88  Chr1    25  C   -   3   2   1   1 between(,11-155)
88  Chr1    88  A   T   7   2   1   1 between(,11-155)
88  Chr1    92  A   C   16  4   1   1 between(,11-155)

This algorithm in python is optimized for very large file using dictionaries

limits = [line.strip().split() for line in open("input.genes")]
limits.pop(0) #remove the header
limits = [map(int,v[1:]) for v in limits]

dict_limits = {}
for start, finish in limits:
  for i in xrange(start, finish+1):
    if i not in dict_limits:
      dict_limits[i] = []
    dict_limits[i].append((start,finish))

OUTPUT = open("my_output.txt", "w")
for reg in open("input.vcf"):
  v_reg = reg.strip().split()
  if int(v_reg[2]) in dict_limits:
    OUTPUT.write(reg.strip() + "\tbetween({})\n".format(str(dict_limits[int(v_reg[2])])))

OUTPUT.close()

you get:

88  Chr1    25  C   -   3   2   1   1   between([(11, 155)])
88  Chr1    88  A   T   7   2   1   1   between([(11, 155)])
88  Chr1    92  A   C   16  4   1   1   between([(11, 155)])

来源：https://stackoverflow.com/questions/30191109/comparing-one-column-value-to-all-columns-in-linux-enviroment

标签

python

Linux

awk

sed

bioinformatics